Abstract
BACKGROUND: Coronary artery calcification (CAC) is a strong predictor of cardiovascular disease, yet metabolism-related molecular alterations underlying CAC remain poorly understood. This study aimed to explore metabolism-related transcriptional changes associated with CAC and identify potential circulating biomarkers for early detection and risk stratification. METHODS: Two transcriptomic datasets, GSE58150 (discovery cohort) and GSE211752 (validation cohort), were analyzed after excluding poor-quality samples. Differential expression, gene set enrichment analysis (GSEA), and metabolic feature identification were performed. Weighted gene co-expression network analysis (WGCNA) identified CAC-related gene modules, whose functions were explored by GSEA. Differentially expressed genes were further evaluated via protein-protein interaction networks and LASSO regression to screen biomarker candidates. Diagnostic value was assessed by ROC curves, and regulatory networks of transcription factors and miRNAs were constructed. A CAC mouse model was then established, with qRT-PCR, Western blotting, and immunohistochemistry used to validate biomarker expression. RESULTS: The discovery cohort included 8 CAC and eight controls; the validation cohort comprised 6 CAC and six controls. A total of 138 genes were upregulated and 104 downregulated in CAC. Functional enrichment indicated alterations in amino acid and vitamin metabolism. WGCNA highlighted the ME2 module as strongly correlated with CAC. GSEA revealed activation of immune-related pathways and suppression of metabolic pathways. LASSO regression identified five biomarker candidates: NRG1, PXDN, ACTL7A, ACSS3, and SHANK3. Among them, PXDN showed the strongest diagnostic performance (AUC 0.95 and 0.83 in discovery and validation cohorts, respectively) and higher expression associated with increased CAC risk. Regulatory analysis suggested PXDN may be modulated by multiple transcription factors and miRNAs. In vivo, PXDN mRNA and protein were significantly elevated in blood and coronary artery tissue of CAC mice. CONCLUSION: CAC is associated with transcriptional alterations in metabolism-related pathways. PXDN was identified as a potential circulating biomarker for CAC, providing a basis for future mechanistic studies and non-invasive risk assessment.